Automated neuroradiological support systems for multiple cerebrovascular disease markers -- A systematic review and meta-analysis

TL;DR

This systematic review comprehensively assesses automated neuroradiology tools capable of identifying multiple cerebrovascular disease markers across CT and MRI. The authors find 29 commercial products and 13 research publications meeting the criteria, with most commercial offerings focused on acute stroke triage (via NCCT/CTA/CTP) and a separate set quantifying WMH and atrophy, typically on MRI. While deep learning dominates the landscape, performance varies across markers and datasets, and there is no validated, open, full-spectrum system covering WMH, ISL, lacunes, PVS, CMB, and atrophy. Key challenges include limited public datasets for all markers, bias and reproducibility concerns, and integration into routine clinical workflows. The study concludes that a comprehensive, automated CSVD support system remains an unmet goal, though DL advances and standardized benchmarking could enable future multi-marker solutions with substantial clinical impact.

Abstract

Cerebrovascular diseases (CVD) can lead to stroke and dementia. Stroke is the second leading cause of death world wide and dementia incidence is increasing by the year. There are several markers of CVD that are visible on brain imaging, including: white matter hyperintensities (WMH), acute and chronic ischaemic stroke lesions (ISL), lacunes, enlarged perivascular spaces (PVS), acute and chronic haemorrhagic lesions, and cerebral microbleeds (CMB). Brain atrophy also occurs in CVD. These markers are important for patient management and intervention, since they indicate elevated risk of future stroke and dementia. We systematically reviewed automated systems designed to support radiologists reporting on these CVD imaging findings. We considered commercially available software and research publications which identify at least two CVD markers. In total, we included 29 commercial products and 13 research publications. Two distinct types of commercial support system were available: those which identify acute stroke lesions (haemorrhagic and ischaemic) from computed tomography (CT) scans, mainly for the purpose of patient triage; and those which measure WMH and atrophy regionally and longitudinally. In research, WMH and ISL were the markers most frequently analysed together, from magnetic resonance imaging (MRI) scans; lacunes and PVS were each targeted only twice and CMB only once. For stroke, commercially available systems largely support the emergency setting, whilst research systems consider also follow-up and routine scans. The systems to quantify WMH and atrophy are focused on neurodegenerative disease support, where these CVD markers are also of significance. There are currently no openly validated systems, commercially, or in research, performing a comprehensive joint analysis of all CVD markers (WMH, ISL, lacunes, PVS, haemorrhagic lesions, CMB, and atrophy).

Figures (4)

• Figure 1:
• Figure 2: Distribution of pathologies targeted by the identified commercial systems.
• Figure 3: Pathology, modality and sample size characteristics of identified research publications.
• Figure 4: Reported Dice similarity coefficients (DSC) for WMH and ISL. Error bars are standard deviation and marker size corresponds to test dataset size. Grey error bars indicated that the standard deviation was not reported and has been estimated visually from box plots as 3/4 of the interquartile range (assumes normal distribution).